An apparatus and method to facilitate finding recommendations for a query image are disclosed herein. A color histogram is determined corresponding to the query image. Determining at least one of a visual pattern included in the query image, a dominant color of the query image, or an orientation histogram corresponding to the query image. Performing comparison of a first image attribute of the query image to a second image attribute of an inventory image corresponding to an inventory item, wherein the first image attribute used in the comparison is selected from among the color histogram, the dominant color, and the orientation histogram. The selection of the first image attribute is based on a confidence score associated with the visual pattern, the dominant color, or a directionality present in the query image.
Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A method comprising: determining, by one or more processors of a machine, a color histogram of a query image; determining, by one or more processors of the machine, a first image attribute of the query image; calculating, by one or more processors of the machine, a confidence score of the first image attribute, the confidence score quantifying a degree to which the first image attribute is present in the query image; and comparing, by one or more processors of the machine, the first image attribute of the query image to a second image attribute of an item image that depicts an item, the comparing of the first and second image attributes being based on the confidence score of the first image attribute of the query image.
A method for recommending items based on images involves these steps: First, determine a color histogram of a query image using a computer processor. Then, determine another image attribute of the query image, such as a visual pattern or dominant color. Calculate a confidence score for this image attribute, indicating how strongly it's present in the query image. Finally, compare this image attribute of the query image to a corresponding image attribute of an item image (an image of an item in a catalog). The comparison is weighted by the confidence score of the query image's attribute, so a higher confidence leads to a stronger influence in the matching process.
2. The method of claim 1 , wherein: the first image attribute is a first visual pattern depicted in the query image; the second image attribute is a second visual pattern depicted in the item image; and the comparing of the first and second image attributes compares the first and second visual patterns based on the confidence score of the first visual pattern.
The item recommendation method described in Claim 1 is specialized to visual patterns. Instead of a generic "image attribute," the method specifically uses a visual pattern identified in the query image as the "first image attribute". The "second image attribute" is a corresponding visual pattern identified in the item image. The comparison step then focuses on comparing these two visual patterns. The confidence score of the visual pattern in the query image determines the importance of that pattern in the comparison against the item image's visual pattern.
3. The method of claim 1 , wherein: the first image attribute is a first dominant color depicted in the query image; the second image attribute is a second dominant color depicted in the item image; and the comparing of the first and second image attributes compares the first and second dominant colors based on the confidence score of the first dominant color.
The item recommendation method described in Claim 1 is specialized to dominant colors. The "first image attribute" becomes the dominant color detected in the query image. The "second image attribute" becomes the dominant color detected in the item image. The comparison step specifically compares the dominant colors of the two images, and the confidence score associated with the dominant color in the query image affects how much weight the query image's color has in the comparison.
4. The method of claim 1 , wherein: the first image attribute is a first orientation histogram of the query image; the second image attribute is a second orientation histogram of the item image; and the comparing of the first and second image attributes compares the first and second orientation histograms based on the confidence score of the first orientation histogram.
The item recommendation method described in Claim 1 is specialized to orientation histograms. The "first image attribute" is an orientation histogram calculated from the query image. The "second image attribute" is an orientation histogram calculated from the item image. The comparison step focuses on comparing these two orientation histograms. The confidence score of the orientation histogram derived from the query image influences the matching process against the item image's orientation histogram.
5. The method of claim 1 , wherein: the comparing of the first and second image attributes is based on the confidence score of the first image attribute transgressing a predefined threshold value that corresponds to the first image attribute.
The item recommendation method described in Claim 1 refines the image attribute comparison by using a confidence threshold. The comparison of the first image attribute of the query image and the second image attribute of the item image only proceeds if the confidence score of the first image attribute exceeds a pre-defined threshold value. The threshold is specific to the type of image attribute being used. This ensures that unreliable attributes don't negatively impact recommendations.
6. The method of claim 1 , wherein: the first image attribute is the color histogram of the query image; and the second image attribute is a color histogram of the item image.
The item recommendation method described in Claim 1 uses color histograms as the primary image attribute. Specifically, the first image attribute determined for the query image is its color histogram, and the second image attribute from the item image is its color histogram. The comparison then directly compares the color histograms of the query image and the item image.
7. The method of claim 6 , wherein: the comparing of the first and second image attributes compares the color histogram of the query image with the color histogram of the item image based on a failure to detect a visual pattern in the query image.
The item recommendation method from Claim 6, which uses color histograms for comparison, incorporates a check for visual patterns. If the system fails to detect a reliable visual pattern in the query image, the comparison relies on comparing the color histogram of the query image with the color histogram of the item image. This provides a fallback mechanism when more specific visual features are absent or unreliable.
8. The method of claim 6 , further comprising: calculating an orientation confidence score based on an orientation histogram of the query image; and wherein the comparing of the first and second image attributes compares the color histogram of the query image with the color histogram of the item image based on the orientation confidence score failing to transgress a predefined threshold value.
The item recommendation method from Claim 6, which relies on color histogram comparisons, also incorporates orientation information. The method calculates an "orientation confidence score" based on the orientation histogram of the query image. If this orientation confidence score falls below a pre-defined threshold, the comparison uses the color histogram of the query image and the color histogram of the item image to find relevant matches. This leverages color when directional information is weak or absent.
9. The method of claim 1 , wherein: the determining of the color histogram of the query image is based on information contained only within a sampling mask area applied to the query image.
In the item recommendation method described in Claim 1, the determination of the color histogram of the query image is based only on a selected area of the image. A "sampling mask area" is applied to the query image, and the color histogram is calculated only from the pixels within this mask. This allows focusing on specific regions of interest within the image, ignoring irrelevant background details.
10. The method of claim 9 , wherein: the sampling mask area defines a central portion of the item image and defines a central portion of the query image.
In the item recommendation method from Claim 9 where a sampling mask is used, the sampling mask is configured to focus on the central region of both images. The mask defines a central portion of the query image and a corresponding central portion of the item image. This focuses the comparison on the main subject within the images, reducing the impact of irrelevant background elements.
11. The method of claim 1 , further comprising: receiving the query image from a device that generated the query image by photographing a subject that exhibits the first image attribute.
The item recommendation method described in Claim 1 includes a step of receiving the query image from a device, such as a smartphone. The device captured the query image by photographing a subject that exhibits the first image attribute. This highlights the use case where a user takes a photo of something they want to find similar items for.
12. The method of claim 1 , further comprising: recommending the item depicted in the item image, in response to reception of the query image and based on the comparing of the first and second image attributes.
The item recommendation method described in Claim 1 concludes with recommending the item depicted in the item image. This recommendation is triggered upon receiving the query image. The recommendation is based on the comparison of the first and second image attributes between the query and item images, suggesting the item to the user as a relevant match.
13. A system comprising: a color histogram module comprising one or more processors of a machine and configured to determine a color histogram of a query image; an image attribute module comprising one or more processors of the machine and configured to: determine a first image attribute of the query image; and calculate a confidence score of the first image attribute, the confidence score quantifying a degree to which the first image attribute is present in the query image; and a comparison module comprising one or more processors of the machine and configured to compare the first image attribute of the query image to a second image attribute of an item image that depicts an item, the comparing of the first and second image attributes being based on the confidence score of the first image attribute of the query image.
A system for recommending items based on image analysis consists of three key modules: A color histogram module that uses processors to determine the color histogram of a query image. An image attribute module, which also uses processors, to determine a first image attribute of the query image (e.g., a visual pattern or color) and calculate a confidence score for that attribute. Finally, a comparison module that compares the first image attribute of the query image to a second image attribute of an item image. The comparison is based on the confidence score calculated for the query image's attribute.
14. The system of claim 13 , wherein: the first image attribute is at least one of a first visual pattern depicted in the query image, a first dominant color of the query image, or a first orientation histogram of the query image; and the second image attribute is at least one of a second visual pattern depicted in the item image, a second dominant color of the item image, or a second orientation histogram of the item image.
The item recommendation system described in Claim 13 is designed to work with different types of image attributes. The "first image attribute" of the query image can be a visual pattern, a dominant color, or an orientation histogram. Correspondingly, the "second image attribute" of the item image can also be a visual pattern, dominant color, or orientation histogram. The system is flexible and can compare different combinations of these attributes depending on what's most relevant.
15. The system of claim 13 , wherein: the comparison module is configured to compare of the first and second image attributes based on the confidence score of the first image attribute transgressing a predefined threshold value that corresponds to the first image attribute.
In the item recommendation system of Claim 13, the comparison module only compares the image attributes if the confidence score of the first image attribute exceeds a predefined threshold value that corresponds to the first image attribute. This thresholding mechanism prevents unreliable attribute data from negatively affecting the quality of item recommendations.
16. The system of claim 13 , further comprising: an image capture module comprising one or more processors of the machine and configured to receive the query image from a device that generated the query image by photographing a subject that exhibits the first image attribute.
The item recommendation system of Claim 13 includes an image capture module that receives the query image from a device. The device captured the query image by photographing a subject that exhibits the first image attribute, e.g. the user taking a photo with their phone.
17. The system of claim 13 , wherein: the comparison module comprising one or more processors of the machine is configured to recommend the item depicted in the item image, in response to reception of the query image and based on the comparing of the first and second image attributes.
The item recommendation system described in Claim 13 is designed to recommend items. The comparison module recommends the item depicted in the item image in response to receiving the query image. This recommendation is determined by comparing the image attributes.
18. A non-transitory machine-readable storage medium comprising instructions that, when executed by processors of a machine, cause the machine to perform operations comprising: determining, by one or more processors of the machine, a color histogram of a query image; determining, by one or more processors of the machine, a first image attribute of the query image; and comparing, by one or more processors of the machine, the first image attribute of the query image to a second image attribute of an item image that depicts an item, the comparing of the first and second image attributes being based on a confidence score of the first image attribute of the query image, the confidence score quantifying a degree to which the first image attribute is present in the query image.
A non-transitory computer-readable medium stores instructions to perform item recommendations. The instructions, when executed by processors, cause the machine to: Determine a color histogram of a query image. Determine another image attribute (e.g., a visual pattern or color) of the query image. Compare the first image attribute of the query image to a corresponding image attribute of an item image. The comparison is based on a confidence score of the first image attribute of the query image, where the confidence score reflects the degree to which that attribute is present in the image.
19. The non-transitory machine-readable storage medium of claim 18 , wherein the operations further comprise: calculating, by one or more processors of the machine, the confidence score of the first image attribute.
The non-transitory computer-readable medium from Claim 18 further contains instructions to calculate the confidence score of the first image attribute. This means the system not only compares attributes but also assesses their reliability.
20. The non-transitory machine readable medium of claim 18 , wherein: the first image attribute is at least one of a first visual pattern depicted in the query image, a first dominant color of the query image, or a first orientation histogram of the query image; and the second image attribute is at least one of a second visual pattern depicted in the item image, a second dominant color of the item image, or a second orientation histogram of the item image.
The non-transitory computer-readable medium described in Claim 18 is flexible in terms of image attributes. The "first image attribute" of the query image can be a visual pattern, a dominant color, or an orientation histogram. Similarly, the "second image attribute" of the item image can also be a visual pattern, dominant color, or orientation histogram. The instructions support working with these different types of image features for the comparison and recommendation process.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
October 22, 2014
March 28, 2017
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.